Falsification of log entries leads to an unusual criticality accident at the Mayak Production Association. During plutonium purification operations, a technician working alone begins transferring plutonium carbonate solution into a holding vessel. He notices a criticality alarm sounding intermittently and exits the room but leaves the transfer going. Soon, the alarm is sounding continuously.

A radiation protection crew arrives. Their instruments measure a flux of 18 R/hour in a corridor near the room. Continued monitoring reveals an average level of 1.5 to 1.8 R/hour with oscillations to tenfold higher levels. It is thought that the oscillations are caused by the transfer pump periodically delivering just enough solution to induce criticality in the holding tank, whereupon a buildup of heat and pressure overcomes the pump and forces some solution back out. Attempts to distribute the solution into separate vessels and stop the oscillating criticality do not succeed until the following day.

Subsequent investigation reveals that a supervisor on a previous shift under-reported the amount of plutonium left in the holding vessel. During the accident and cleanup, five individuals receive doses of between 0.24 and 2.0 rem. Although the task of cleanup and recovery of plutonium from the apparatus is arduous, no contamination results and the equipment suffers no damage.

After operating for two years, the 3-MW SL-1 reactor at the National Reactor Testing Station in Idaho has been down for routine maintenance over the Christmas 1960 holidays. The three-man crew on duty the night of 3 January is preparing it for full-power operation the following day. Apparently, for reasons unclear, one operator withdraws the central control rod from the core as rapidly as possible. The excursion reaches a peak power of 20,000 MW and produces 130 Megajoules of energy. The resulting steam explosion destroys the reactor, lifting the entire core 9 feet before dropping it back in place. Two of the operators are killed instantly; the third is critically injured. Cooling water spreads intensely radioactive fission products throughout the building. Some contamination escapes into the local environment. Most of it settles within a 3-acre zone outside the building, but iodine-131 levels 100 times normal are found in vegetation 20 miles from the site.

Entering the building is extremely hazardous due to radiation levels of 500 to 1,000 R/hour. Search for the three victims is done by site personnel working in relays. The survivor is removed promptly; he dies a few hours later of massive head injuries. Due to higher priorities, the other two bodies are left in place for a few days. (One of those two is found on the ceiling of the reactor room, impaled there by the control rod.) Investigation and cleanup are similarly hampered; but within a few months the reactor is dismantled and its pieces safely buried on site.

A B-52 develops a leak in an in-wing fuel tank. A fire erupts, causing the plane to break up in midair and crash 12 miles north of Seymour Johnson AFB in Goldsboro, NC. The bomber's two 24-MT Mark 39 hydrogen bombs are released. One of the bombs deploys its parachute and lands relatively undamaged. The other plunges into a muddy field and disintegrates. Three of the plane's 8 crewmen are killed in the initial explosion; the other five parachute to safety.

Investigators find three of the intact bomb's four arming devices have activated. Only the pilot's safe/arm switch has prevented a detonation. Much of the other bomb is recovered, but the thermonuclear stage has penetrated to an estimated depth of 180 feet, and groundwater flooding prevents excavation. The Air Force purchases the plot of land and fences it off, monitoring periodically for contamination. None is reported.

CPP-601, the Idaho Chemical Processing Plant's main building, is five days into renewed operations following a year-long shutdown. In an evaporator system, a concentrated solution of enriched uranium is forced back up into the reservoir from which it flowed. The reason is unclear; a bubble of high-pressure air is suspected. The resulting criticality sets off alarms and all personnel evacuate the facility, with exposures of only 60 mrem or less. The criticality ends a few minutes after it began at 0950. Work is allowed to resume at 1445.

Continuous air monitors are the only instruments from which any data can be obtained, and their data are not of much use for determining either intensity history or total energy. Thus the parameters of the event are poorly known; but yield is estimated as 6x10¹⁷ fissions. Evaluation suggests that causes of the accident include unfamiliarity of personnel with equipment after the long shutdown, and the poor operating condition of some equipment.

K-19, a Soviet Hotel-class submarine, suffers a major accident off Norway after a leak in a pipe causes its reactor cooling system to fail. Reactor core temperatures reach 800°C, nearly hot enough to melt the fuel elements. But the crew, working in a dangerously radioactive compartment, is able to prevent this by rigging an emergency cooling system. The accident contaminates the crew, parts of the ship, and some of the ballistic missiles it carries. Eight of the crew die from acute radiation sickness before help can arrive. When it does arrive, the surviving crew transfer to another submarine and the K-19 is towed back to the Kola Peninsula.

Through a complex sequence of events, a significant quantity of U-235 (in the form of UF6) accumulates in the oil reservoir of a vacuum pump driving the gas purifier of the number 6 desublimation-sublimation stage (DSS-6) of the gaseous diffusion plant at the Siberian Chemical Combine. The oil is normally changed each 15 days because the UF6 lowers the efficiency of the vacuum pump. On this date, equipment breakdowns elsewhere in the diffusion cascade cause DSS-6 to work harder than normal. A high-radiation alarm activates at 0445. Per procedure, the operator shuts the pump down and summons the radiation control officer. The RCO cannot locate the source of radiation. He does not check the pump room, basically because he dismisses it as a possibility. Radiation flux in the control room measures about 9 mR/h and decreasing. He authorizes resumption of operations.

At 0730 the operator turns on the pump. The criticality alarm sounds almost immediately and the operator sees a flash of Cerenkov radiation. He turns off the pump and runs to a telephone to report. This time, the RCO finds 36 mR/h at the exterior of the DSS-6 building, He sends the operator to the hospital, where his exposure is estimated as 200 rad. He suffers mild radiation sickness and recovers. Investigation discloses the causes of the first and the much larger second excursion. They are too complex to describe here.

In mid-October, a USAF Jupiter mobile intermediate-range ballistic missile (IRBM) is struck by lightning at its launch site near the Gioia del Colle Air Base in Italy. Three other nearby missiles are struck on different dates through August 1962. In each case, the missile's thermal batteries are activated, and on two occasions tritium-deuterium gas is injected into the "pits' of the 1.4-MT warheads, partially arming them.

After the fourth strike, the Air Force places lightning diversion towers at all Jupiter missile sites in Italy and Turkey. (It is noteworthy that the deployment of these IRBMs prompts the USSR to place similar missiles in Cuba, bringing on the "Cuban Missile Crisis" of October 1962.)

A power transient occurs in a 75kg mass of highly enriched (93%) uranium metal moderated by paraffin. This experiment is the last of a series at Oak Ridge measuring reactivity changes when increments of paraffin or U-235 are added to the system. All previous configurations have been subcritical when fully assembled. However, this one goes critical while the halves of the assembly are still 2.7 inches apart. Automatic safeties scram the system within 50 mSec.

The fission yield is between 10¹⁵ and 10¹⁶ fissions. This range is confirmed by the fact that the paraffin does not melt and the surface of the uranium is unmarred. No significant personnel exposures result, and the laboratory returns to normal use within 1.5 hours.

Sometime in 1961, a Soviet nuclear ballistic-missile submarine of early design (probably a Hotel class) is operating near the coast of England when a coolant pipe breaks. The level of radiation is reportedly 5R/h at the break. The flooding coolant spreads radiation to several parts of the ship, including its missile tubes. Crew members are seriously contaminated. The ship is able to return to port where, after a two-month ventilation, its missiles are transferred to two diesel-powered submarines for their test launches.

Though it is not credited, this is clearly a CIA report. The date is indicative.

In the Recuplex system process plant at the Hanford Works, operator error apparently contributes to a later excursion in a plutonium solution. The solution overflows a receiver tank and winds up on the floor of a solvent extraction hood and in the sump. The operator, contrary to orders, opens a valve that allows this spilled material to be lifted into a transfer tank. Later addition of more plutonium solution and subsequent moderation following mixing or de-aeration of the tank cause the excursion. An initial spike of 10¹⁶ fissions is followed by 37.5 hours of steadily decreasing emission, for a total yield of 8x10¹⁷ fissions. As it is a Saturday morning, only 22 people are in the plant. They all evacuate promptly when the criticality alarm sounds. Three receive significant doses of 110, 43 and 19 rem.

Witness testimony does not fully jibe with the technical findings about this accident, so the above reconstruction should be regarded as the best plausible explanation. The accident causes no damage to equipment and no contamination release. However, it does bring about the final shutdown of the plant. (This has already been planned, since a new production facility is approved.) The first use of an instrumented and television-equipped remotely-controlled vehicle occurs during the investigation of this accident.

The second Thor launch in Operation Dominic is called Starfish. It ends like the first Bluegill test. The missile malfunctions and is destroyed by the RSO at 65 seconds. Its nuclear warhead, a W-49 type rated at 1.45 MT, does not detonate but is destroyed by its high-explosive triggers. Debris including plutonium contaminates both Johnston Island and Sand Atoll.

Note: The Starfish Prime test on 9 July is successful. However, this detonation at 248 miles altitude is notorious for EMP effects which cause electrical outages on Oahu.

This accident occurs at the Mayak Production Association in a facility where plutonium feedstocks are purified and cast into ingots. The residues from this operation include slag that contains recoverable quantities of plutonium. Recovery involves using nitric acid to dissolve the plutonium metal in the residue and performing other chemical processing.

The cause is the usual one of inadvertently allowing too much solution in a single container. The criticality triggers an alarm and all personnel evacuate. Operating valves, heaters and stirrers remotely from the shift supervisor's control room, operators bring the situation under control in about an hour. During this time two other excursions occur, the third being the largest. It is discovered that 1,324 grams of plutonium, both solution and precipitate, have accumulated in the vessel. Dose rates measured at the glove box containing this vessel average 2R/h. Total yield is estimated as 2x10¹⁷ fissions. Worker exposures are negligible. Some solution is forced onto the floor of the glove box, but no equipment damage occurs and cleanup takes only a short time.

Bluegill Double Prime is the third failure of a Bluegill test and the fourth malfunction of a Thor IRBM in Operation Dominic. In this case, the RSO destroys the missile at 156 seconds. Its nuclear warhead does not detonate but is destroyed by its high-explosive triggers. Some plutonium contamination of Johnston Island occurs.

Note: Bluegill Triple Prime is a success on 26 October. The missile carries its warhead to 31 miles altitude, where it detonates.

A train of the Louisville & Nashville Railroad derails near Marietta, GA while carrying nuclear weapons components. These are not damaged, but three couriers are injured in the derailment.

Note: At a December 1962 AEC symposium in Germantown, MD, a report is presented that lists 47 accidents involving nuclear materials. The report calls 17 of them "serious".

Mistakes by two different operators on different shifts are the cause of this accident in a uranium metal production building at the Siberian Chemical Combine. The process involved is similar to that cited in Incident 095 for plutonium recovery: dissolving the waste in nitric acid, followed by further chemical processing.

The first error is made because the facility uses two different formats for recording uranium mass per batch: as a mass fraction (that is, the uranium-to-precipitate mass ratio) or as grams of uranium per kilogram of precipitate. The latter is the format used where the accident occurs. Two waste containers are received for processing. The attached analysis forms record the uranium mass fraction of 0.18. The shift supervisor transcribes this as 18g/kg on his work orders, instead of 180. Referring to these work orders, an operator loads 2 kg from one container and 5kg from the other into a processing tank. Thus he is processing ten times more uranium than he thinks he has in his tubes and tanks. The actual amount is 1,260 grams of Uranium. He begins dissolving the precipitate with acid. Then a shift change occurs.

The oncoming operator completes the dissolution. Per normal procedures, he takes a sample and sends it to the lab for a concentration check. Later he telephones for the result. This should reveal the original error. Unfortunately, by coincidence, he's given the result for another sample — one with a ten times lower concentration.

Next, the supervisor for that shift decides to recycle the solution for the next batch of precipitate. This too is normal procedure. The same transcription error is in effect, and another 1,225g of uranium are added. Thus the vessel holds over 2.5kg of uranium, an amount close to critical mass for its geometry.

Ultimately, criticality does occur. Four people standing 10m from the vessel receive doses of 6 to 17 rad before they evacuate in response to the alarm. Over the next ten hours, eight more excursions take place in the vessel. The cycle is broken on 31 January by draining part of the solution to another container. No damage to the vessel occurs and there is no contamination of the area. Based on lanthanum-140 in samples taken afterward, the total yield is 7.9x10¹⁷ fissions.

Researchers at Lawrence Livermore Laboratory are doing multiplication measurements on various configurations of uranium and reflectors. As is common, the reflector is mounted to a stationary support and a lift table is used to raise the uranium sample into proximity with it. The apparatus is located in a heavily shield vault.

This particular experiment involves moving the uranium closer in steps, making a measurement each time. Seven steps have been done with complete safety. On the eighth, the system unexpectedly goes highly supercritical. An explosive sound is heard, the scram activates, and alarms sound. After a few seconds, the uranium can be seen melting and burning. About 15kg of uranium burn. Another 10kg melt and spread over the floor. The reactor room is heavily contaminated.

No dose to anyone in the building exceeds 0.12rem. The yield is later measured as 3.67x10¹⁷ fissions. The cause of the excursion is thought to be a mechanical misalignment: The central cylinder is slightly off-center. When raised into the reflector, it carries the rings upward with it — until the eighth step, when the cylinder shifts back toward centerline and the rings fall down around it.

The nuclear attack submarine USS Thresher (SSN-593) has just been refitted and is undergoing sea trials in the Atlantic about 220 miles east of Boston. During a deep dive, the vessel sinks in 2,600 meters of water with loss of all 129 hands. It is thought that the silver-solder joint of a pipe fails, resulting in uncontrollable flooding of the vessel's interior. Such joints have failed during destructive testing in the shipyard.

Six pieces of the Thresher are later photographed on the sea bottom. The reactor compartment remains intact and no radiation leakage from it has been detected.

A vacuum system in a facility at the Siberian Chemical Combine is protected by traps. Operations inevitably introduce droplets of solutions into the system. The corrosive chemicals involved in reprocessing can attack the components of the vacuum system, and any solutions that collect within it represent loss of product. To prevent large accumulations within the traps, they are fitted with level sensors that trigger drain valves as needed. These sensors operate when the solutions touch pairs of electrodes, allowing a current to flow. They are designed for processes that use water solutions, which always conduct electrical current. Unfortunately, the facility is doing a mix of processes: some use water solutions, while others use organic solutions which are at best weakly conductive.

First-stage traps in this particular setup typically fill to the trigger point four times per day. They then drain some liquid into a larger, second-stage trap. When that fills to the preset level, it is drained back into the processing stream. The first-stage traps contain some aqueous solution, but also an organic solution of lower density, which floats on top of the aqueous. Thus, draining is triggered only when the level of aqueous rises high enough.

The criticality occurs in the large trap, which has not drained in 8 days. An alarm sounds and employees are evacuated from the area. The trap is found to hold a large quantity of organic solution containing enriched uranium. Over the next 16 hours, a series of relatively weak excursions takes place at intervals. Total yield is 6x10¹⁶ fissions. The accident is ended by manually siphoning the liquids into smaller reservoirs. Fortunately, this vessel is located in a corridor not frequented by employees. The largest dose to anyone is under 5 rem. The accident results in replacement of the traps by an improved system and in redesign of some processes.

A B-52D is carrying two nuclear weapons from Westover AFB in Chicopee Falls, MA to its home base at Turner AFB in Albany, GA. This is a tactical ferry flight, meaning that the weapons are merely stored aboard the aircraft for transport; they are not mounted and connected as droppable bombs.

About halfway through its flight, the bomber encounters violent turbulence at 29,500 feet. It attempts to climb to better conditions at 33,000 feet, but encounters worse turbulence and suffers structural failure. It crashes 17 miles southwest of Cumberland, MD. The two bombs are recovered "relatively intact".

A chemical plant at Wood River Junction, RI is designed to recover highly enriched uranium from scrap material left over from reactor fuel element production. This is the United Nuclear Fuels Recovery Plant. It begins operations on 16 March 1964. The plant receives the scrap as uranyl nitrate solution, concentrates it, purifies it, and finally bubbles it through a tank of the solvent trichloroethane (TCE) to remove other organic solvents used in processing. A single charge of TCE is expected to last 6 months or more; but experience shows that the limit is one week, due to the rate at which it picks up uranium.

A procedure is developed to reclaim the uranium (still a very low concentration of 400 to 800 ppm) from the TCE before discarding it by washing it with sodium carbonate solution. Originally this is done manually by shaking the mixture in 11-liter bottles. On 16 July, though, a large amount of TCE has accumulated. An operator is given permission to begin washing it in a large stationary vessel. This variation is taught to an operator on a different shift, and by 24 July the two operators have washed 10 to 12 bottles each. The sodium carbonate solution remains in the large vessel, its uranium concentration slowly rising.

Concurrently, someone else dissolves a plug of uranium nitrate crystals out of an evaporator hose, putting the concentrated solution in bottles identical to those used in the TCE wash. These bottles ARE labeled as highly concentrated. Nevertheless, one of the wash operators mistakes one for a bottle of contaminated TCE and pours it into the large vessel to wash it. Immediately there is a flash, the radiation alarm sounds, and about 20% of the solution splashes onto the walls, the ceiling and the operator. He gets up and runs to an emergency building.

Ninety minutes later, the plant superintendent and a shift supervisor enter the building. The superintendent removes the bottle, still upended over the tank, and turns off the tank stirrer. Unknown to these men, since the radiation alarm is still sounding, a second excursion occurs when the tank's liquid slumps into a new geometry. This is discovered when their doses, 100 rem and 60 rem respectively, are found to be much higher than expected. The original excursion imparts a 10,000 rad dose to the operator; he dies 49 hours later. Total yield is 1.30±0.25x10¹⁷ fissions. The equipment is undamaged; the room is decontaminated.

An accelerator accident in Illinois produces local exposures ranging from 29,000 to 240,000 rad. One individual loses an arm and a leg.

I'm guessing this occurred at FermiLab in Batavia, Illinois. Then again, it might be Argonne National Laboratory. Both are located near Chicago. Or it could be a private facility elsewhere in the state.

The staff at Building 242 of the Electrostal Machine Building Plant are changing over their uranium oxide production process from 2% to 6.5% enriched uranium. The line shuts down on 19 October and everything is thoroughly cleaned over the next three days. The accident occurs when flakes of uranium oxide enter the vacuum system and collect in a vacuum supply vessel. The alarm sounds at 11:10 and Building 242 is immediately evacuated. The facility's chief physicist enters 50 minutes later and measures a gamma exposure rate of 3.6R/h at 1.5m from the vacuum supply vessel. Fission yield falls between 5x10¹⁵ and 10¹⁶. One person may have received a dose of 3.4 rads.

Careful recovery operations are taken to divide up the mass of uranium. It is determined to hold 4.6kg of U-235. Investigators determine that sometime after production resumes on 22 October, the primary filter goes missing from the vacuum duct, and the secondary filter is improperly fastened. These filters are supposed to be checked once per shift. Also, each shift is required to take a water sample from the vacuum system and have it checked for uranium concentration. This has not been done since production resumed.

The aircraft carrier USS Ticonderoga is traveling from station off Viet Nam to its base at Yokosura, Japan. Somewhere between these two locations, an A4-E Skyhawk strike fighter carrying a B-43 hydrogen bomb rolls off the carrier's elevator and plunges into the sea at a point where it is 16,200 feet deep. Neither the bomb, the plane, nor its pilot is ever recovered.

One source claims there is concern at the Pentagon that water pressure at that extreme depth will cause the bomb to detonate, and adds that whether or not an explosion did occur is unknown. Neither seems very likely.

More substantial concerns drive the dispute over exactly where this incident occurs. Japan's law forbids nuclear weapons on its territory, including U.S. bases on that territory, so admitting the Ticonderoga carries such weapons would underscore a violation of U.S. military agreements. It would also be an admission that we deployed nukes to Viet Nam.

Lackadaisical record keeping is the cause of a criticality accident at the Mayak Production Facility It occurs in the residue recovery area of a metal and fissile solution processing building. On the day before the accident, a supervisor directs an operator to process batch 1726 in equipment intended for material with less than 1% uranium content. This is a rules violation, since 1726 contains more than 1%. After the process completes, 1726 is transferred to another glovebox which already holds multiple batches of residue. It is analyzed and its 44% by-weight concentration is noted in the log but not conveyed to the central data registry.

Another operator, set to begin the next processing step, notices that the figure for batch 1726 is missing from its accountability card and calls the laboratory to obtain it. He is given by mistake the result for batch 1826: a concentration 138 times smaller. The second operator properly records this on the card attached to batch 1726.

The next day, during dissolution of the residue, these errors result in a criticality alarm. The operator leaves the area and goes to the central control room. There, alarms begin sounding for progressively distant sensors. The vessel responsible is soon identified; it is supporting multiple excursions and emitting an average level of 8R/h at a distance of 2m. The usual recovery procedures are judged too time-consuming for this case. Emergency personnel develop an alternate plan, each step of which will be done by a different person and take no more than 30 or 60 seconds:

1. Remove two gloves from the glovebox to gain access.
2. Unlock and open the feed hopper lid.
3. Drop a crumpled-up ball of cadmium foil into the solution to poison the fission reaction.

This plan works. No one receives a dose greater than 0.3 rad. The next day, the solution is transferred into safe containers and sent to a special facility. It is later reprocessed after the cadmium is removed. Total yield of the event is 5.5x10¹⁷ fissions, from 11 excursions. The facility is back in operation several days later.

The VENUS reactor is a tank-type, water-moderated, critical assembly machine in a laboratory at Mol, Belgium. It is fueled by 1,200kg of uranium oxide with an enrichment of 7%. Fission regulation is done by 2 control rods and 8 safety rods, all remotely controlled by an operator. An additional 8 rods are provided; these must be manually adjusted and locked down.

On this date the operator decides to change over to a new rod pattern. By means of written instructions, he directs a technician to insert a new manual rod and remove another. The rod configuration is such that the reactor should be subcritical by 1 safety rod, 2 control rods and 1 manual rod. (The last control rod is being inserted into the core by its mechanism.) The operator overlooks the requirement that the water should be drained before any manual rods are adjusted. The technician does not wait for the slow insertion of the final control rod; he also reverses the sequence in his instructions, removing one manual rod before inserting the other. The reactor goes critical at this point. He leaves the room immediately on seeing the flash, but already has received a severe dose, primarily of gamma rays. The energy release is 4.3x10¹⁷ fissions — not enough to cause damage or even a steam explosion in this particular reactor. Because of his position leaning over the assembly, the technician's exposure is highly uneven. Estimates are that his head receives 300 to 400 rem, his chest 500, his left ankle 1,750, and the end of his left foot 4,000. He is treated in hospital and survives, but his left foot does not.

During in-flight refueling near Palomares, Spain, a B-52 collides with a KC-135 tanker aircraft. Eight of the 11 crewmen aboard the two planes are killed. The B-52 splits in two, and the KC-135 goes down in flames, its 40,000 gallons of extra fuel contributing to the blaze. Wreckage rains down on an area of land and water 100 square miles in extent. One of the B-52's four hydrogen bombs, having deployed its parachute, lands intact near Palomares. Two others fall freely; their HE charges explode on impact, scattering radioactivity over nearby farms. The fourth bomb plunges into the Mediterranean Sea 12 miles off the coast.

During the ensuing cleanup, 1,400 tons of lightly contaminated soil and vegetation are dug up and shipped to a nuclear waste dump at Aiken, South Carolina for disposal. The U.S. settles claims by 522 Palomares residents for $600,000, and gives the town a $200,000 desalination plant.

The weapon lost in the Mediterranean sets off one of the largest search and recovery operations in history. It involves 3,000 Navy personnel and 33 vessels, not counting the ships, planes and people providing logistics and other support. (One source reports a total of 12,000 people.) The submersible Alvin locates the bomb after two weeks, but it is not recovered until 7 April. The whole effort takes 80 days, and is later fictionalized in the film Men of Honor.

Using a very highly enriched solution of uranyl nitrate (97.6% U-235), researchers at Oak Ridge are conducting experiments to determine the critical concentration of solution in a thin aluminum sphere with a thick water reflector. The sphere is connected by a flexible hose to a cylindrical reservoir. Solution height in the sphere is adjusted by raising or lowering the cylinder.

After a measurement, an attempt to pull the sphere back subcritical by lowering the cylinder is ineffective. Upon inspection, an air bubble is seen in the hose. Enough solution is drained into a third container to bring the sphere subcritical. Then the cylinder is jostled in an attempt to remove the bubble. The system suddenly goes critical again. It is thought that the air bubble broke free after the jostling ended, pushing enough solution into the sphere to drive it critical.

The excursion expels approximately 90 ml of solution out of the sphere and onto nearby equipment. The yield is found to be 1.1x10¹⁶ fissions. The modest cleanup required is accomplished promptly. A simple modification fixes the problem of air bubbles in the connecting hose.

Chelyabinsk-70 is part of the Russian Federal Nuclear Center (VNIITF), located in the southern Ural mountains between the cities of Etakarinberg and Chelyabinsk. Intensive work is under way there to develop powerful reactor-based neutron sources for radiation-tolerance studies.

Two nuclear criticality specialists are working overtime on a Friday evening to complete a second assembly on a FKBN lift mechanism. This is to be a duplicate of one constructed during the day — with one exception: a polyethylene sphere is to fill the central space which in the earlier assembly is empty. While the senior specialist uses a handheld control to lower the upper half of the reflector, the junior specialist stands by to guide it into place. The accident occurs as the reflector is about to make contact with the core. There are no criticality alarms installed at this time, but a scram system immediately drops the core, terminating the reaction.

Both men remain conscious and retain their self-control. They are able to report the accident and to call for an ambulance. The senior specialist carries out dose estimates for himself and his assistant and writes the details of the event in his personal log. The senior specialist's dose is in the 5-10 Sv range; the junior gets 20-40 Sv. Both are taken to the local hospital and immediately flown to the Bio-Physics Institute in Moscow. The junior specialist dies three days later; the senior specialist survives for 54 days.

Investigators conclude that, in addition to underestimating the effect of the polyethylene sphere, the senior specialist violated several rules, and that both men were guilty of overconfidence and excessive haste.

K-129, a diesel-powered Soviet submarine of the Golf-II class, sinks in 16,000 feet of water 750 miles northwest of Oahu. The vessel takes 80 sailors with it to the bottom, as well as three SS-N-5 ballistic missiles and (probably) two nuclear torpedoes. The cause of the sinking is thought to be an internal explosion, perhaps due to hydrogen from the ship's storage batteries. (The Soviets asserted that the USS Swordfish (SSN-579) was the cause; some Russians still believe this. The Swordfish was shadowing K-129, and it did undergo secret nighttime repairs at Okinawa shortly after K-129's loss.)

The U.S. Navy's SOSUS underwater listening system gives them the area where K-129 sank. In 1969, the exact location of the wreck is discovered by the USS Mizar (T-AK-272). It is then surveyed by USS Halibut (SSGN-587) and the bathyscape Trieste II. The CIA begins Project Jennifer to recover the submarine for analysis, and commissions a mining company owned by Howard Hughes to build a vessel that can raise it. On 12 August 1974, Hughes' Glomar Explorer brings the hulk halfway home, but it breaks apart. The engine room and torpedo rooms are lost, and only the center section is recovered. The ballistic missiles too reportedly fall back to the seabed.

Thus, the CIA missed out on the "good parts"; but what was retained is still classified. Indeed, much mystery remains. High-level sources hint at an informal agreement not to tell the full story of the sinking, or that of the USS Scorpion (SSN-589) in May 1968.

Note: The date of this event is somewhat ambiguous. Sources claim 8 March or 11 April 1968, and a Russian Web page says 1967.

The nuclear-powered attack submarine USS Scorpion (SSN-589) is returning from a three-month training exercise in the Mediterranean Sea to its base at Norfolk, VA. It sinks 400 to 500 miles southwest of the Azores, taking with it 99 sailors and two Mark 45 ASTOR nuclear torpedoes.

The wreckage is later photographed at its 10,000-foot depth by the research vessel Mizar (T-AK-272). No evidence of sabotage or foul play is found, laying to rest suspicions that the Soviets are somehow involved in the sinking. The Navy continues to monitor the wreck for leakage from its reactor. None has been found to date.

The Soviet submarine K-27, a modified November-class equipped with two VT-1 type reactors, is undergoing sea trials. Reactor power suddenly drops, and the crew are unable to restore normal power output. At the same time, gamma radiation levels in the reactor compartment rise to 150R/h and radioactive gases vent into it from the safety buffer tank. The reactor is shut down and the crew are taken off the boat.

Investigation shows that 20% of the core has melted due to a problem in the reactor's liquid-metal cooling loop. The heavily contaminated submarine is deemed beyond economical repair. It is scuttled in a shallow area of the Kara Sea in 1981.

a) The Cass City Chronicle reports that "two months ago" a plant in Pennsylvania found 220 lbs of enriched uranium unaccounted for. Another source claims the missing uranium wound up in Israel.

b) "In addition, the Central Intelligence Agency (CIA) has long believed that during the mid-1960s, Jerusalem diverted more than 100 kilograms (220 pounds) of highly enriched uranium, enough for four rudimentary weapons, from a private U.S. processing plant at Apollo, Pennsylvania. In 1968, 200 metric tons (220 short tons) of uranium concentrate disappeared while in transit on the Mediterranean Sea from Antwerp, Belgium, to Genoa, Italy. American and European officials are convinced that the material — essential for the long-term operation of the Dimona reactor — wound up in Israel."

A fire in a glove box at Rocky Flats consumes 5kg of plutonium, releasing significant contamination onto the grounds of the facility. Hundreds of railway cars are used to transport the contaminated soil to Idaho Falls, where it is left in unlined trenches.

The Colorado Committee for Environmental Information deploys a team of scientists with sophisticated instruments, putting officials on notice that the public now has the capability to discover and report releases of radiological contamination. The Committee's work in response to the fire provides evidence of a years-long buildup of wastes on the grounds of Rocky Flats.